Strategy to Validate Sensor-Placement Methodologies in the Context of Sparse Measurement in Complex
This IEEE Sensors Journal article (May 15, 2020) by Bertola, Costa, and Smith proposes a validation strategy for sensor-placement methodologies when only sparse field measurements are available. It pairs predicted information-gain metrics (information entropy and joint entropy from a hierarchical sensor-placement algorithm, and signal-to-noise ratios) with observed performance measured as falsification rates using Error-Domain Model Falsification (EDMF). The paper presents a stepwise validation procedure, statistical tests (rank correlation Γ, linear regression, R^2, p-value threshold α=0.05), and three full-scale bridge case studies (Singapore Flyover, Exeter Bascule Bridge, Rockingham Bridge) to demonstrate that the hierarchical (joint-entropy) algorithm predictions correlate well with observed falsification performance while a simple SNR-based strategy performs poorly in at least one case.
Questions this document answers
10.1109/JSEN.2020.2969470
IEEE Sensors Journal, Vol. 20, No. 10, May 15, 2020
Numa J. Bertola, Alberto Costa, and Ian F. C. Smith
0.05
Singapore Flyover; Exeter Bascule Bridge (UK); Rockingham Bridge (Australia)